[XPM]unofficial jhPrimeminer thread

[mumus]

I'm solo mining xpm with 10 computers using jhprimeminer -xpm option. However, all of them are connected to one server with Primecoin client.
I just realized that the all the computers are solving the same blocks.
New Block: 332966
The numbers are all the same across all the computers, and I just got an orphan block.
Are they solving different parts of that block, or are they solving the same thing redundantly, producing an orphan block? I hope it's the first case. Other wise, I'm just wasting the energy.

One way make sure that all the miners has unique work is to specify different addresses in the -xpm parameter.

[1714135420]

1714135420

[1714135420]

1714135420

[1714135420]

1714135420

[1714135420]

1714135420

[ssl303]

Thanks for the info. I've been using the same coin address. Does this mean I've been wasting energy/time?
Do I need to create new account for each computer, or can I just use getnewaddress on the same account to get another coin address?
If I need to create another account, do I need to put that information on the primecoin.conf as well?
rpcuser=account1
rpcpassword=password
rpcuser=account2
rpcpassword=password
rpcuser=account3
rpcpassword=password
...

[mumus]

Thanks for the info. I've been using the same coin address. Does this mean I've been wasting energy/time?
Do I need to create new account for each computer, or can I just use getnewaddress on the same account to get another coin address?
If I need to create another account, do I need to put that information on the primecoin.conf as well?
rpcuser=account1
rpcpassword=password
rpcuser=account2
rpcpassword=password
rpcuser=account3
rpcpassword=password
...

About your first question, not sure. Depends on computer time also.
You only need to use the getnewaddress. No need for separate accounts.

[ssl303]

Thanks again,
After I generated new address with getnewaddress on the same account, I changed the address that comes after -xpm option for each computer. However, the number that comes after "New Block:" are still the same across the all the computers. Is each computer supposed to solve different block number?
Also, I notice that it takes a long time (a day or more) for jhprimeminer to catch up to 10ch even 9ch every time I restart the process. Does it mean whenever I restart the process, I'm wasting time for jhprimeminer to catch up? Does jhprimeminer need to solve lower ch in order to get to higher ch, or is there a chance for jhprimeminer to solve 10ch before lower ch? .

[mumus]

Thanks again,
After I generated new address with getnewaddress on the same account, I changed the address that comes after -xpm option for each computer. However, the number that comes after "New Block:" are still the same across the all the computers. Is each computer supposed to solve different block number?
Also, I notice that it takes a long time (a day or more) for jhprimeminer to catch up to 10ch even 9ch every time I restart the process. Does it mean whenever I restart the process, I'm wasting time for jhprimeminer to catch up? Does jhprimeminer need to solve lower ch in order to get to higher ch, or is there a chance for jhprimeminer to solve 10ch before lower ch? .

The block number should be the same for all miners (for everybody in the primecoin network). What is important is that the internal algorithm is working on a different hash to find prime chains.
Regarding to your other questions I would suggest to read a little about the algorithm behind the miner from the document that was published by the creator of primecoin http://primecoin.org/static/primecoin-paper.pdf
When the miner is running you are not wasting time. Lower ch (prime chain) is a byproduct of searching for the right chain length set by the network (difficulty).

[ssl303]

Thanks a lot for explaining this!

[btcxyzzz]

Please add HTTP proxy support for miner.

+1 !!

[muto]

What miner is currently the best for solo mining?

[primer-]

Warning to all miners using Rdebourbon's 4.0F2 miner - he has never released source code and its impossible to tell if some blocks found are being siphoned off to his xpm wallet address.

[gh2k]

Warning to all miners using Rdebourbon's 4.0F2 miner - he has never released source code and its impossible to tell if some blocks found are being siphoned off to his xpm wallet address.

This concerns me as well... However, would it not be possible to find this out by looking at the daemon's logs?

[primer-]

Warning to all miners using Rdebourbon's 4.0F2 miner - he has never released source code and its impossible to tell if some blocks found are being siphoned off to his xpm wallet address.

This concerns me as well... However, would it not be possible to find this out by looking at the daemon's logs?

No, its not possible to trace through debug.log. Dont waste your time with Rdebourbon's 4.0F2 miner, use hp11 instead!

[jrovins]

Is there any chance that versions from rdebourbon and tandyuk get merged so both windows and linux version get updated at the same time?

Yes, actually.. This is on the cards.. tandy now has write access to my repository on github, so I expect that as we go forward the two forks will merge into one..

Hi,
I really wanted to run a private server, with a few Linux workers behind it. I got tired of waiting, so I decided to waste a few days of my holiday vacation, and become a merge monkey, I pulled down tandyuk's and rdebourbon's latest from git hub, and with the help of smartgit, and meld, I eventually got something that compiled.  I have been running it for about a week now, and it is fairly stable with the primecoin-qt wallet as a server, and has mined a few blocks. But it crashes with primecoind as a server.

I consider this an alpha test version, the -xpm option is functional.

There are some features that are not working, and there was the following commit, in tandyuk's master branch that I am now sure that I should have backed out:

-------------------------------------------------------------
commit 5a5f8f58534f1bbed0c2810268e799d6f23b3305
Author: James Tandy <james@jamestandy.com>
Date:   Sat Sep 14 23:56:30 2013 +0100

    Commit for other devs to take a look,
    
    I dont recommend trying to use this build for mining!
--------------------------------------------------------------

This one removed the automatic cache tuning from the linux version, among other things. I am thinking about re-doing the merge again without this commit.  (if I have the time )

Guidance from tandyuk or rdebourbon would be welcome.

Merges are on github  for access & review:

https://github.com/jrovins/jhprimeminer

Note: this is an experimental version, no binaries available, build from the source, and use at your own risk.

[ssl303]

According to jhprimeminer, one of the computers has 1 on 10ch, but it doesn't appear on transactions list. How long does it take? It's been more than a day.

[primer-]

According to jhprimeminer, one of the computers has 1 on 10ch, but it doesn't appear on transactions list. How long does it take? It's been more than a day.

It should be instant if the its valid. Make sure its higher than the current difficulty.

[ssl303]

That's strange. It says it found 1 block at 10ch which is current difficulty. I use Mumus 8.2.

[John122]

Does anybody have a working link for a good 32-bit jhPrimeminer version?
I just got hold of an old laptop and all the links for 32-bit versions seem to be down...

Also, primer-, I can't seem to find a link for that hp11 version (be it 32 or 64-bit) you're talking about. Could you link me to it?

[ssl303]

Another computer found a block, and it's not getting credited. I'm going to stop using Mumus build.

[cal_guy]

Another computer found a block, and it's not getting credited. I'm going to stop using Mumus build.

If I'm not mistaken at difficulties above 10 a 10-ch no longer has a guaranteed chance of solving a block.

[primer-]

Does anybody have a working link for a good 32-bit jhPrimeminer version?
I just got hold of an old laptop and all the links for 32-bit versions seem to be down...

Also, primer-, I can't seem to find a link for that hp11 version (be it 32 or 64-bit) you're talking about. Could you link me to it?

Get the updated hp11 source directly from github. Public release is expected any day now (hp12)
git clone https://github.com/mikaelh2/primecoin/

[ssl303]

If I'm not mistaken at difficulties above 10 a 10-ch no longer has a guaranteed chance of solving a block.

What do you mean? Right now the difficulty is at 10.22630489.
If jhprimeminer says "10-chain count:    1," aren't I supposed to get credited for 9.55 xpm?